Medium-term speaker states—A review on intoxication, sleepiness and the first challenge

Schuller, Björn; Steidl, Stefan; Batliner, Anton; Schiel, Florian; Krajewski, Jarek; Weninger, Felix; Eyben, Florian

doi:10.1016/j.csl.2012.12.002

Cited by 50 publications

(42 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In particular, the bulk of studies found that the rate of speech and the overall amplitude decrease after alcohol consumption, whereas the pitch variability, the mean fundamental frequency (F0), and the sentence duration increase [1]. Likewise, sleepy speech manifests changes in prosody (e. g., monotonic intonation, shifted speech rate, reduced syllable duration due to retarded cognitive speech planning), articulation (e. g., slurred pronunciation, speech errors, disfluency), and voice quality (e. g., nasal or breathy speech) [15]. Specifically, in accordance with the findings of alcohol intoxication, the sleepiness-induced changes of speech parameters include a general decrease in speech rate [16] and an increased average absolute deviation of intensity [17].…”

Section: Introductionmentioning

confidence: 99%

Cross-Domain Classification of Drowsiness in Speech: The Case of Alcohol Intoxication and Sleep Deprivation

2017

Self Cite

View full text Add to dashboard Cite

In this work, we study the drowsy state of a speaker, induced by alcohol intoxication or sleep deprivation. In particular, we investigate the coherence between the two pivotal causes of drowsiness, as featured in the Intoxication and Sleepiness tasks of the INTERSPEECH Speaker State Challenge. In this way, we aim to exploit the interrelations between these different, yet highly correlated speaker states, which need to be reliably recognised in safety and security critical environments. To this end, we perform cross-domain classification of alcohol intoxication and sleepiness, thus leveraging the acoustic similarities of these speech phenomena for transfer learning. Further, we conducted in-depth feature analysis to quantitatively assess the task relatedness and to determine the most relevant features for both tasks. To test our methods in realistic contexts, we use the Alcohol Language Corpus and the Sleepy Language Corpus containing in total 60 hours of genuine intoxicated and sleepy speech. In the result, cross-domain classification combined with feature selection yields up to 60.3 % unweighted average recall, which is significantly above-chance (50 %) and highly notable given the mismatch in the training and validation data. Finally, we show that an effective, general drowsiness classifier can be obtained by aggregating the training data from both domains.

show abstract

Section: Introductionmentioning

confidence: 99%

Cross-Domain Classification of Drowsiness in Speech: The Case of Alcohol Intoxication and Sleep Deprivation

2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…This feature obtained a very competitive RMSE of 10.17, well below the Challenge test audio baseline of 14.12. Given the consistent performance of this feature in system development and in other paralinguistic tasks [24], [25], this result is not surprising. The second audio system chosen was another KL-mean system.…”

Section: Test Set Resultsmentioning

confidence: 53%

“…Mel Frequency Cepstral Features (MFCC) are one of the strongest performing spectral features, when combined with Gaussian Mixture Models (GMM), for classifying either low/high levels of depression [22], [23], or the presence/absence of depression [7], [8]. Classification using MFCCs in combination with GMM-UBM (universal background model) supervectors has recently gained popularity for many paralinguistic tasks: many entrants to similar Interspeech Challenges on speaker affect, intoxication and sleepiness have used this style of system to obtain competitive results [24], [25]. Motivated by recent results showing a decrease in energy variability with increasing levels of depression, due in part to a decrease in the motor action associated with speech production [17], [20], and by results suggesting that this decrease in variability can be captured in a GMM [22], we explore a range of GMM-UBM supervector systems in combination with Support Vector Regression (SVR) for the task of predicting depression.…”

Section: Non Linguistic Audio Cuesmentioning

confidence: 99%

Diagnosis of depression by behavioural signals

Cummins

Joshi

Dhall

et al. 2013

Proceedings of the 3rd ACM International Workshop on Audio/Visual Emotion Challenge

104

View full text Add to dashboard Cite

Quantifying behavioural changes in depression using affective computing techniques is the first step in developing an objective diagnostic aid, with clinical utility, for clinical depression. As part of the AVEC 2013 Challenge, we present a multimodal approach for the Depression Sub-Challenge using a GMM-UBM system with three different kernels for the audio subsystem and Space Time Interest Points in a Bag-of-Words approach for the vision subsystem. These are then fused at the feature level to form the combined AV system. Key results include the strong performance of acoustic audio features and the bag-of-words visual features in predicting an individual's level of depression using regression. Interestingly, in the context of the small amount of literature on the subject, is that our feature level multimodal fusion technique is able to outperform both the audio and visual challenge baselines.

show abstract

“…For a more detailed description of ALC, see Schiel et al (2012). A subset of ALC was used for the INTERSPEECH 2011 Speaker State Challenge (Schuller et al, 2012).…”

Section: Speech Data Used In Experimentsmentioning

confidence: 99%

The influence of alcoholic intoxication on the short-time energy function of speech

Schiel

2014

The Journal of the Acoustical Society of America

Self Cite

View full text Add to dashboard Cite

This study investigates rhythmic features based on the short-time energy function of speech signals with the aim of finding robust, speaker-independent features that indicate speaker intoxication. Data from the German Alcohol Language Corpus, which comprises read, spontaneous, and command&control speech uttered by 162 speakers of both genders and various age groups when sober and intoxicated, were analyzed. Energy contours are compared directly (Root Mean Squared Error, statistical correlation, or the Euclidean distance in the spectral space of the contour) and by parameterization of the contour using the Discrete Cosine Transform (DCT) and the first and second moments of the lower DCT spectrum. Contours are also analyzed by Principal Components Analysis aiming at fundamental "eigen contour" changes that might encode intoxication. Energy contours differ significantly with intoxication in terms of distance measures, the second and fourth DCT coefficients, and the first and second moments of the lower DCT spectrum. Principal Components Analysis did not yield interpretable "eigen contours" that could be used in distinguishing intoxicated from sober contours.

show abstract

Medium-term speaker states—A review on intoxication, sleepiness and the first challenge

Cited by 50 publications

References 36 publications

Cross-Domain Classification of Drowsiness in Speech: The Case of Alcohol Intoxication and Sleep Deprivation

Cross-Domain Classification of Drowsiness in Speech: The Case of Alcohol Intoxication and Sleep Deprivation

Diagnosis of depression by behavioural signals

The influence of alcoholic intoxication on the short-time energy function of speech

Contact Info

Product

Resources

About